Home of Professional Cinematography since 1996

The Human Eye

Published : 6th October 2003


Does anyone know the equivalent ASA and f-stop range of the human eye?
Does the eye have some sort of frame rate, sampling rate or refresh rate?

D.A. Oldis
Director/Cinematographer
Winston-Salem, NC



D. A. Oldis writes:

>Does anyone know the equivalent ASA and f-stop range of the human >eye?
>Does the eye have some sort of frame rate, sampling rate or refresh >rate?


Also, is it possible to adjust the eye's refresh rate by going into the control panel?

Bob Patrick
NY/NJ DP



> Does anyone know the equivalent ASA and f-stop range of the human >eye?
> Does the eye have some sort of frame rate, sampling rate or refresh >rate?


The human eye can readily adapt to light levels from at least .03 F/C (full moon, no other illumination) for scotopic vision (rods), to 15,000 F/C (bright daylight in snow) for photopic vision (cones).

I'm not sure how that would translate into ASA and f-stop.

Detectable refresh rate (temporal frequency) varies between 40 and 80 hz (depending on the lighting ratio of source to ambient light). The eye can detect flickering (temporal separation) of equal intensity light sources at intervals as low as 15-20 ms.

Finally, resolution is between 50 and 60 cyc/deg. in bright light, but that number has a lot of qualifiers that I won't go into. (Our hearing is limited to 20-20K hz, but the brain can detect missing harmonics at much higher frequencies.)

Jessica Gallant
Los Angeles based Director of Photography
West Coast Systems Administrator, Cinematography Mailing List
http://www.cinematography.net



>I'm not sure how that would translate into ASA and f-stop.


Jessica quotes a sensitivity range of 500,000:1, or 19 stops. That comes from two factors -

1/. Aperture. The iris opens and shuts, but only over a limited range - approximately 4 stops, or a ratio of 16:1. (That's approximate - of course the eye focuses by changing its focal length, which in turn affects the aperture).

2/. Adaptation, effected by the sensitivity of the rods and cones. This is the closest equivalent to the ASA range of the eye - in film terms. I learnt that this was up to 100,000 to 1, but even 30,000 to 1, or 15 stops (to fit Jessica's figures) is quite impressive.

On top of this of course is the sensitivity range of the retina at any given aperture and adaptation. It's quite a lot more than film, with a _linear_range of 9 stops or so. However, there is a massively extended toe and shoulder, and at extremely bright or dark brightness levels, the eye still has some slightly reduced discrimination. It is estimated that the total potential response range of the eye in all conditions is at least 10^10 to

1. (That’s 10 billion to 1). Maybe 10^12:1.

Why do you ask, D.A. Oldis? Planning to build one?

Dominic Case
Group Technology & Services Manager
The Atlab Group

http://www.atlab.com.au



Dear Sir!

What is relationship between human eye and 24fps ?

Thank You.

N. Prakash



Human eye vision is around 60 fps as per our tests of Showscan proved years
ago.

Jim Dickson



D.A. Oldis wrote :


>Does anyone know the equivalent ASA and f-stop range of the human >eye?

Dominic Case wrote :

>It is estimated that the total potential response range of the eye is at >least 10^10 to 1. Why do you ask, D.A. Oldis? Planning to build one?

Actually I'm trying to figure out what the human eye considers a "normal" exposure. And at what point during the day or at what light level does the eye reach full aperture and begin to interpret ambient light as "darkness."

This bizarre query has occurred to me several times over the years. Most recently I thought about it on a 35mm TV spot that required some mini-DV shots at dusk. Prior to rolling tape the DV camera was set to auto iris. As the sun started to set I noticed the camera's gain slowly kicking in, keeping the image in the viewfinder at what the camera felt was a "normal" exposure.

It didn't take long before the image in the viewfinder was significantly brighter than the actual location looked to my eye. In fact, in the viewfinder it didn't look dark at all. It looked like noon on an overcast day. The only thing that gave it away as being night was the fact that street lights and car headlights were turned on and blowing out.

Now this didn't come as a revelation to me. I've seen it many times before when using video cameras, and for the actual shots we simply stop down the lens. But for some reason I was fascinated by the fact a cheap little DV camera (cheap, that is, compared to my 35mm gear) could have greater sensitivity than my eyes.

Likewise I'm always intrigued after shooting night scenes on film. When I start editing the transferred night time footage I love the look of the flash frames at the head and tail of each shot. Those few frames have an incredibly surreal look; it may have been midnight when we shot, but the overexposed flash frames look like they were filmed on the surface of the sun or on Earth during a nuclear explosion. It's clearly night time. But it's really bright. I've tried in vain for years to figure out a way to get that look for an entire scene without resorting to time lapse photography.

If anyone has any ideas on how to do it, I'm all ears!

Anyhow, the point is that I was reminded of something I've always known but rarely think about: that given enough exposure time, film is much more sensitive than the human eye. So when you say that the human eye has a response range of 10,000,000,000 to 1, that's pretty impressive. But it's nowhere near the potential response range of film because with film you have infinite control over the exposure time, whereas the exposure time for the eye is, as I understand it, fixed.

Thanks,

D.A. Oldis
Director/Cinematographer
Facade Films
Winston-Salem, NC



>cheap little DV camera…could have greater sensitivity than my eyes.

Now are we talking sensitivity, or adaptability? Different things.

>The only thing that gave it away as being night


That's the clue. The eye can, as we've said, adapt over an enormous range. So, to a lesser but still useful extent, can film or video photography. But it's what the brain does with the image that is significant. We can see quite clearly by day or at dusk, and without any external clues, we wouldn't be aware of what the absolute light levels are. But the brain is very smart at picking up the clues: as you say, street lights coming on but also the relative brightness of sun and sky to land, the presence and length of shadows, the colour temperature of the light, all tell the brain how to interpret what it sees.

The relative signals from colour-sensitive cones, and more-sensitive but colour-blind rods in the retina also play a part. A night-time image appears less colour-saturated because the monochromatic rods supply a relatively stronger image. Subconsciously we calibrate the brightness of the ambience from that. Your video or film camera will record the same degree of saturation in dim light, which makes night look more like day than it does to the naked eye.

I guess we may also get feedback from the iris setting that the eye has achieved (it's harder to focus if your eyesight is failing a little).

>with film you have infinite control over the exposure time,

Yee..es, and that's how you get the flash-frame effect you mentioned, or time-exposure stills of night-for-day. But it's not infinite. There's something called reciprocity failure, which says that for very short or very long exposures you get a limiting return. In other words, two minutes isn't twice as much exposure as one minute. The silver halides spend so long waiting around for the next photon to arrive that they "forget" how much they've seen already.

Whether the eye has a variable time-exposure property, or whether it is purely an intensity-based response is a bit more complex. Let's not open that book this time.

>I'm all ears!

All eyes would be more useful

Regards

Dominic Case
Group Technology & Services Manager
The Atlab Group



>at what light level does the eye reach full aperture and begin to interpret >ambient light as "darkness."

Trying to compare any form of shooting to the human eye is like comparing a trip to the store for milk to a trip to the moon. Nothing does, can, or ever will come anywhere near the human eye in terms of the way it sees, and more importantly how the brain interprets what it sees.

Walter Graff